add gpu ci (#56)

.github/copy-pr-bot.yaml

@@ -0,0 +1,4 @@
+# Configuration file for `copy-pr-bot` GitHub App
+# https://docs.gha-runners.nvidia.com/apps/copy-pr-bot/
+
+enabled: true

.github/workflows/ci.gpu.yml

@@ -0,0 +1,72 @@
+name: CI (GPU)
+
+on:
+  push:
+    branches:
+      - main
+      - "pull-request/[0-9]+"
+
+concurrency:
+  group: ${{ github.workflow }}-on-${{ github.event_name }}-from-${{ github.ref_name }}
+  cancel-in-progress: true
+
+jobs:
+
+  build-gpu:
+    if: github.repository == 'mhaseeb123/nvstdpar'
+    name: GPU (${{ matrix.name }}, CUDA ${{ matrix.cuda }})
+    strategy:
+      fail-fast: false
+      matrix:
+        include:
+          - { name: "nvc++ 23.3", cuda: "12.0", cxx: "nvc++",  tag: "nvhpc23.3",       gpu: "v100", sm: "70", driver: "latest", arch: "amd64" }
+    runs-on: linux-${{ matrix.arch }}-gpu-${{ matrix.gpu }}-${{ matrix.driver }}-1
+    container:
+      options: -u root
+      image: rapidsai/devcontainers:23.06-cpp-${{ matrix.tag }}-ubuntu22.04
+      env:
+        CUDA_VERSION: "${{ matrix.cuda }}"
+        NVIDIA_VISIBLE_DEVICES: "${{ env.NVIDIA_VISIBLE_DEVICES }}"
+    steps:
+      - name: Run nvidia-smi to make sure GPU is working
+        run: nvidia-smi
+      - name: Checkout nvstdpar 
+        uses: actions/checkout@v3
+        with:
+          persist-credentials: false
+      - name: Build and test GPU schedulers
+        shell: bash -elxo pipefail {0}
+        env:
+          cxx: "${{ matrix.cxx }}"
+        run: |
+          mkdir -p /workspaces;
+          cp -ar . /workspaces/nvstdpar;
+          cd /workspaces/nvstdpar;
+          # Configure
+          cmake \
+            -DCMAKE_BUILD_TYPE=Debug \
+            -DSTDPAR=gpu \
+            -DCMAKE_CXX_COMPILER="$cxx" \
+            ..
+          # Compile
+          make
+          # cmake --build build -v;
+          
+          # Print sccache stats
+          sccache -s
+          
+          # Tests
+          # ctest --test-dir build --verbose --output-on-failure --timeout 60;
+          # Examples
+          # ./build/examples/nvexec/maxwell_cpu_st --iterations=1000 --N=512 --run-cpp --run-inline-scheduler
+          # ./build/examples/nvexec/maxwell_cpu_mt --iterations=1000 --N=512 --run-std --run-stdpar --run-thread-pool-scheduler
+          # ./build/examples/nvexec/maxwell_gpu_s --iterations=1000 --N=512 --run-cuda --run-stdpar --run-stream-scheduler
+
+
+  ci-gpu:
+    runs-on: ubuntu-latest
+    name: CI (GPU)
+    needs:
+      - build-gpu
+    steps:
+      - run: echo "CI (GPU) success"